Roger H. Farrell
Ph.D. (1959) University of Illinois
Mathematical statistics, measure theory
Retired as of July 1, 1999, I am still semi-active in the department. I am not active in research.
My research concerned the application of decision theory methods to statistical problems to try and characterize completely good and bad methods of estimation and testing. Useful decision theory methods can involve development of inequalities, compactification of spaces, and study of the way sequences of measures converge.
Proof of a necessary and sufficient condition for admissibility in discrete multivariate problems (with L. D. Brown), J. Mult. Annal. 24 (1988), 46–52.
All admissible linear estimators of the vector of gamma state parameters with application to random effects models (with W. Klonecki and S. Zontek), Ann. Statist. 17 (1989), 268–281.
A lower bound for the risk in estimating the value of a probability density (with L. D. Brown), Jour. Amer. Statist. Assoc. 85 (1990), 1147–1153.
Estimations of accuracy in testing (with J. T. G. Hwang, G. Casella, C. Robert and M. T. Wells), Ann. Statist. 20 (1992), 490–509.
Spitzer and Bohnenblust, revisited (1997), preprint.